99% of AI Products Will Fail. Here's Why.
The internet is drowning in "AI slop." Here's how to stop building generic tools and start building defensible moats.
I’ve been thinking a lot about what AI can do, and more specifically, how valuable it is or isn’t.
The more I think about it, the less excited I become about the AI industry as a whole. Subsequently, the more excited I become about the high-level AI workflows we can create.
Today, I want to share some fascinating statistics with you, and then contradict myself by explaining where the exciting AI opportunities actually are.
Let’s get started.
The AI Slop Is Real And No One Is Benefiting
A friend of mine recently reminded me of “the dead internet theory.” The notion behind this idea is that when the entire internet is a network of AI agents creating content, talking to each other, and interacting with each other, then the internet has lost its life.
I am seeing this play out firsthand.
For example, I have a big LinkedIn following. As recently as a year ago, I could publish a post and get a few dozen comments from legitimate followers.
Now, every comment is AI. My “engagement” is higher than ever, but my conversion rate from engagement to newsletter subscribers (the one you’re reading now) is lower than ever. Because no real people are engaging with my content.
AI talking to AI.
Yesterday, I read two articles that completely blew my mind.
Graphite, which is a well-known technical SEO company, released a study where they discovered that over half of the text-based content on the internet is written by machines.
In addition, an independent group of AWS researchers concluded that more than 57% of all text-based content (including social media, forums, and chatrooms) is created by AI.
The dead internet is here.
But this phenomenon isn’t just limited to blog posts and social media comments. Just as AI commoditized our words, it is now rapidly commoditizing our code. What’s weirder, is that AI is coming to kill SaaS as well.
The Commoditization Of Code: Why Vibe Coding Will Kill Commoditized SaaS Products
I’ve been thinking a lot about how AI creates a race to the bottom.
What got me thinking about this was my personal brand. A few weeks ago, I started recording daily videos and publishing them. These videos are not AI, it’s me. My own thoughts, my own perspectives.
These videos are doing AWESOME! I am building a big following, and I don’t use AI anywhere. (The same is true for this newsletter btw. Humanity for the win).
So then I started thinking about AI software products. Every day, a new web application software is released and new product releases are added into already existing platforms. It’s easier than ever to use platforms like Episolo to build your own web app in hours, if not minutes.
So if it’s so easy, how come these new web apps aren’t thriving?
Conversely, how come my own AI product, Directorly, is thriving? We are generating new sales and making more money every day with new sign-ups.
The answer is specialization.
Directorly is a custom-coded AI tool. It’s not something you can prop up in an afternoon with a generic prompt, because the functionality is hyper-specific to people who want to build robust and sophisticated directories. We built a defensible moat by deeply understanding the exact data structures and proprietary workflows our users actually need. You can not build directorly with an out-of-the-box AI generator.
Specialization is the key to standing out online with your content. It’s also the key to building products and using AI to increase efficiencies in your current business.
Generic AI Tools Are The SaaS Version Of AI Slop
If you want to make money with AI, you need to build something special.
Specific knowledge, technical skills, and industry specific insights are more important than basic coding skills and app development.
99% of all AI products will fail, because it’s a race to the bottom. Weirdly enough, the people who choose to use AI as a tool, and build agent models to solve hyper-specific problems while using hyper-specific data sets, are the ones who will win.
The more personalized and specific, the better.
AI will not make it easier to make money. In a lot of ways, I think AI will make it harder to make money, because if you want to stand out in a marketplace where everything becomes an instantly commodified version of itself, then you have to live in a place where you bring specific and unique viewpoints into a market that is not easy to learn or understand.
Does anyone really care about another AI-generated task management system? Or macro tracker? Or content generator?
No. I don’t think they do.
As always, the riches are in the niches.
In a world of instant generation, the real winners will be the entrepreneurs willing to invest the heavy time and money required to build deep solutions that AI cannot do alone. They are the ones who will own the future.
Thanks for reading.
Love you guys. Talk to you on Friday with the next issue of Healthcare Business.
Tim
Scale Your Business With A Predictable, Automated AI Agent System
Book a call and we’ll provide you with a free workflow audit and automation strategy. No catch. No commitments.
You will get a proven, done-for-you system to streamline your operations and reduce overhead costs on demand.
You’ll have access to …
custom-built AI agents tailored specifically to your unique business workflows
end-to-end infrastructure management (we host and maintain the AI so your team doesn’t have to)
real-time performance tracking straight to your phone, so you can check your KPIs in plain English via text or Slack
ongoing operational reviews to continuously find and automate new bottlenecks as you scale
If you’re ready to level up your operations without adding headcount, then click the button below and fill out the form.
Let’s automate your business together.
Speed to market is super fast thanks to AI but also made defending what you built if it gets traction impossible. But specialization alone doesn’t save you either. A hyper-specific product nobody finds is just an expensive hobby. Distribution is what makes or breaks it, niche or mass market, AI or not. Always has been.